Fluoro Derivatives of Ethane and Ethylene. IV - Journal of the

May 1, 2002 - George V. D. Tiers , Harvey A. Brown , Thomas S. Reid. Journal of the American Chemical Society 1953 75 (23), 5978-5979. Abstract | PDF ...
0 downloads 0 Views 319KB Size
June, 1936

FLUORO DERIVATIVES OF ETHANE AND ETHYLENE

S87

Cyclohexyl Derivatives.-Cyclohexyl bromide treated r d u x e s t o the reaction vessel. The heating is t o be done in an oil-bath, and should be as high as 180' a t the end of with HgF, in a solvent reacts a t once, a t room temperathe operation t o ensure yields better than SO%, usually ture. The reaction mass becomes purple, an indication about 90%. In this manner CH2BrCHBrF and CHP- that cyclohexylene is formed. Since it is impractical to BrCHFg have been prepared from CHnBrCHBrg; simi- separate the cyclohexyl fluoride from the reaction mass, larly CHBrlCHBra yields quantitatively CHBrgCHFBr or settling is allowed to take place and the mercuric salts are CHBrtCHFz, according to the proportion of the reactants. filtered off. The clear filtrate is shaken with water, and Acyl Fluorides.-The mode of operation is similar t o examination of the water layer indicates the presence of that recommended for ethyl fluoride. Mercuric fluoride hydrogen fluoride and cyclohexyl alcohol in quantities reacts vigorously with acetyl chloride, and the acetyl corresponding t o about 10% yields of cyclohexyl fluoride. fluoride thus formed volatilizes through the reflux con- The same kind of results have been obtained from orthodenser. To gage the yields, acetyl fluoride was received and from para-dibromocyclohexane. in ethanol, with which it reacts at once. The quantity of Summary ethyl acetate recovered corresponded to about a 50% yield of acetyl fluoride; it is felt that this yield could be considerMercuric fluoride is proposed as a new fluorinatably improved. Reaction with Ethanol.-A small amount of gas is ing agent. Its preparation, properties and uses formed, and the reaction stops promptly. The gas is a are described. mixture of ethylene and ethyl fluoride, the latter in about TgE MIDGLEY FOUNDATION 2% yield. RECEIVED JANUARY 30, 1936 COLUMBUS, OHIO

[CONTRIBUTION FROM THE

DEPARTMENT OF CHEMISTRY AT THE

OHIO STATE UNIVRRSITY]

Fluoro Derivatives of Ethane and Ethylene. IV BY ALBERTL. HE^ AND MARY W. RENOLL Preceding papers' have described fluoro derivatives of ethane and ethylene containing no hydrogen, one hydrogen and two asymmetrically placed hydrogen atoms, The present paper presents fluoro derivatives containing two asymmetrically placed hydrogen atoms, and places emphasis on the different courses followed by the fluorination, when different fluorinating agents are used. Swarts2 has shown that the fluorination of acetylene tetrabromide by means of antimony fluoride yields successively CHBr2CHBrF and CHBr2CHF2, where the fluorination stops completely and that acetylene tetrachloride yields CHC12CHClFand CHC12CHF2,together with a very small amount of CHClFCHF2. Swarts' observations and the physical constants reported by him have been verified, and there is no doubt that his determination of the molecular structure of the difluorides is correct. The only point which was never duplicated was the production of small quantities of CHCLFCHF2, and this may be due to the fact that Swarts operated in platinum equipment; it is known that platinum has a catalytic action on fluorination. The experiments carried out in this Laboratory were performed in steel, copper or nickel. (1) THISJOURNAL, 56, 1726 (1934);68,402,404(1936). (21 Swarts, M e m . Acad. Roy. Sci. Bela., 61, 1-94 (1901).

Since the fluorination of acetylene tetrabromide is considerably more rapid than that of acetylene tetrachloride, an effort was made to fluorinate CHClBrCHClBr, in the hope that this would yield CHClFCHClF, a compound which should undergo further fluorination without great difficulty.a However, the usual treatment with antimony fluoride gave rise to the following compounds: CHC12CHClF, CHC12CHF2, CHC4CC4F and CHC12CCIF2, which were produced in small quantity, as by-products, and have been previously described, and two new compounds, CHClBrCHClF, and CHC1BrCHF2. The characteristics of these new compounds were measured and computed as in the preceding papers.' Compound B. p., "C. A t mm. d20d

n2'D

MRD c1, % Br, %

F,% Mol. wt.

EXPERIMENTAL VALUES CHClBrCHClF CHClBrCHF2 124.7-125.1 82.3-82.5 736 743 1.932 1.879 1.4776 1.4173 28.48 24.04 36.1 19.7 40.9 44.6 9.6 21.0 194.2 178.0

Neither substance freezes a t solid carbon dioxide temperature. By subtracting from the observed (3) See THISJOURNAL, 58,882 (1936).

888

w.RENOLL

ALBERTL.HENNEAND MARY

MRD the increments given by "I. e. T."for C, H, C1 and Br, the atomic refraction of F is computed as 1.08 in the case of the difluoride, and 0.65 in the case of the monofluoride. The value 1.08 is excellent. The value 0.05 is low, but by no means out of line with observations on similar compounds. For instance, Swarts observes F = 0.83 in CHCl&HCIF, while in the case of CHBrzCHBrF, the computed value of F becomes even negative. As it was now evident that fluorination was invariably proceeding in an asymmetrical fashion, efforts were made to cause further fluorination by combining the following experimental conditions : (1) using either CHC12CHFz or CHBrCHF2 as starting material; (2) adding increasing amounts of SbFsCl2 as catalyst; (3) operating under pressure; and (4) raising the temperature to 150-160". In all cases considerable decomposition took place, but by operating in large quantities, it became possible to collect and isolate some reaction products: typical runs on 2 kilos of GHzClzFz gave 350 g. boiling below 30', while 1 kilo of C2H,Br2Fz gave 200 g. boiling below 50". Repeated refractionation gave, in addition to many compounds resulting from the replacement of one or both hydrogen atoms by chlorine or bromine, the following results. From the chlorides: a compound boiling a t 6.1", which proved to be CHaCICFsand no indication of anything boiling about 17", where CHClFCHF2 boils. From the bromides : some substance boiling at 23-25 " , where CH2BrCFsshould boil, and some substance boiling about 40°, where CHBrFCHFz should boil; but in neither case, any indication of a compound a t -23" or -35", where CHFzCHFz and CHzFCFa would boil, respectively. To demonstrate the course of the rearrangement, the compound boiling a t 6" was prepared in larger quantity, and purified. Its characteristics were found t o be: boiling point 6.1'; do, 1.389; %OD 1.3090; MRD obsd. 16.4; C1 found 30.4010, calcd. 30.0%; F found 48.0%, calcd. 48.5%; mol. wt. found 118, calcd. 118.5. Computing the atomic refraction of fluorine as above gives F' = 1.1, a good value. These properties indicate a gross formula C,H2C1FB. Chlorination in sunlight gives exclusively CClaCFs, a known compound, and thus demonstrates that the formula of the original compound is CH2ClCFa. As the corresponding bromide could not be obtained in a good state of purity without unjustified d o r t s ,

VOl. 5 s

its formula was accepted by analogy based on boiling points; in general fluorobromides boil about 20' higher than the corresponding fluorochlorides. The resources of fluorination with antimony fluoride having thus been exhausted, another fluorinating agent was used, namely, mercuric fluoride, and its use made it possible to obtain the following sequence CHBrzCHBrz -+CHBr2CHBrF -+ CHBrzCHF2+CHBrFCHF, + CHFzCHF2. At all stages a single compound was made and the complete absence of isomerization made it a simple task to purify the products. The mono- and the difluoride have been adequately described by S w a r k 2 The tri- and the tetrafluoride are new substances, and their characteristics are as follows: CHFBrCHFz boils a t 4041" under 735 mm., has a density dlo4 1.874, refractive index n l " ~1.36175, from which the molecular refraction (observed) is computed as 19.13. Subtracting from this observed MRD the sum of the increments for C, H and Br, an atomic refraction of 1.08 is found for fluorine, which is correct. It contains 49.0% Br and 34.8% F (calcd. 49.06 and 34.96) and its molecular weight is 161.5 (calcd. 103). CHFzCHF2 boils at -23", has an observed molecular weight of 102 (computed, 102) and contains 74.2% F (74.4% calcd.). The adoption of the proposed formula in preference to CHzFCFa, the only other possibility, is based on the facts that the treatment with HgF2 gave a single trifluoride, and a single tetrafluoride, and that the boiling points of their isomers, namely, CHZBrCFa and CHaFCF3, would have been about 25" and -36", respectively, both sufficieotly lower to eliminate any ambiguity.

Experimental HgFs and CHBrzCHFz.-One hundred cc. (1.08 mols) of CHBreCHFa and 100 g. (0.42 mol) of HgFn were heated electrically in a horizontal nickel tube connected to a vertical steel pipe acting as reflux condenser and dephlegmator, and closed by a needle valve; a pressure gage and thermocouple recorded the conditions of the system. The reaction products were led through a calcium carbonate suspension and a drying tube, into a receiver cooled with solid carbon dioxide. The pressure was kept between 120150 pounds by purging and the temperature was held at 180-190°. Reaction continued for about nine hours. By slow purging it was possible to avoid removal of much trifluoride. After final purification 16 cc. of CHFzCHFz and 23 g. of CHFtCHFBr were obtained; the remainder was unreacted CHBrzCHF2.

FLUOROETHANES AND FLUOROETHYLENES

June, 1936

HgFz and CHBr2CHBr2. CHBr2CHBr2 reacts with HgFz already a t 100", very easily at I50-1fiflo, to form CHBrzCHFl quantitatively. ~

Summary The course of the fluorination of CHXzCHXz

889

compounds is presented. The following new compounds CHClBrCHC1F, C H C ~ B ~ C H C F ~H,~ CICF3, CHFBrCHFz and CHFZCHFZare described. COLUMBUS, OHIO

RECEIVED MARCH23, 1936

[ CONTRIBIJTION FROM THE DEPARTMENT O F CHEMISTRY AT THE OHIO STATE UNIVRRSITY]

Fluoroethanes and Fluoroethylenes. V BY ALBERTL. HENNEAND MARYW. RENOLL The present paper presents the fluorination of halo derivatives of ethane containing three or four atoms of hydrogen in their molecule, and describes several new fluorinated compounds. Preceding papers' have shown the course of the fluorination in derivatives containing less hydrogen and have emphasized the tendency to lose hydrogen halides, to rearrange and to resist fluorination with antimony fluorides. The main result of the present investigation is the finding that, when the halogen atoms are all located on the same carbon atom, fluorination proceeds rapidly to completion, without simultaneous chlorination or loss of hydrogen halides. In marked opposition, derivatives having both hydrogen and an halogen on each of their carbon atoms resist fluorination with antimony fluorides considerably, and when they are affected they undergo important decomposition and chlorination. The fluorinated compounds exhibit a marked difference in their stability, in favor of the asymmetrical compounds: CH3CHFzis perfectly stable, while CHzFCHzF decomposes spontaneously, and CHBFCHFBhydrolyzes slowly while CH3CF3 is nearly inert physiologically as well as chemically. Two methods of fluorination have been used, namely, (1) the interaction of an organic polychloride with antimony fluoride in the presence of SbC&or Br, and (2) the interaction of a polybromide with mercuric fluoride. The first procedure gave excellent results with CH3CC13,fair results with CH3CHCl2, mediocre results with CHzCHCIz, and no results with CHzCICHzCland with CH3CHzCl. The second method was used with success to fluorinate CHZBrCHClBr, CHZBrCHBr2, CHdCHFz, CH2BrCHzBr,CH3CHBr2and CH3CHzBr. (1)

T H I S JOURNAL,

68, 402, 404, 887 (1938).

The properties of the new compounds only are listed in Table I. The known compounds produced had physical properties in good agreement with those published by Swarts (see Beilstein). The analyses of the new derivatives were successfully performed as previously described, while the molecular weights were measured by freezing point depression in benzene for the liquids and by quantitative analysis of known volumes for the gases. All were close to the theoretical quantities, and appear in Table I. The atomic refractions for fluorine listed in the seventh column of Table I were obtained by subtracting from the observed molecular refraction the sum of the increments for C, H, C1 and Br. Most values are in line with those obtained by Swarts3 for similar compounds, but the very first one is entirely out of line, and this remains unexplained. This value was repeatedly verified, and no emor was detected in the experimentation; it is true that CH3CC12F is not extremely stable, but its decomposition is quite slow and does not explain the fact that the density is about 0.04 lower than could be expected by analogy. The chlorine analysis is also too high, an indication of impurity. The molecular weight is correct. Experimental Details (1) Fluorination of CH&C&.-The fluorinating agent is a mixture of 90% SbFa and 10% SbFsClz. I t is placed in a metal container, equipped with a metal dephlegmator bearing a pressure gage, thermometer and needle valve. The container is cooled in an ice-bath, with the needle valve closed. This creates a partial vacuum, and makes it possible to suck ice-cold CHsCCL into the equipment. The relative quantities of reagents depend on whether the mono-, di- or trifluoride is the desired final product. The reaction vessel is allowed to come slowly to room temperature, as the reaction starts very vigorously. By control (2) Hubbard and Henne, THIS JOURNAL, 66,1078(1934) (3) Swarts, J. chim. p h y s . , 20, 30 (1923).